Australia looks to US for AI regulation

Labor is looking to Joe Biden’s new artificial intelligence reforms as a blueprint it could follow to regulate the technology.

Industry Minister Ed Husic says Washington’s move to sign an executive order governing the safe and responsible use of AI this week could help inform Australia’s response to the unregulated technology. Picture: NCA NewsWire / Martin Ollman​

The federal government is looking to US President Joe Biden’s new artificial intelligence reforms as a blueprint it could follow to regulate the emerging technology, as Labor prepares its own sweeping regulations to strengthen domestic rules.

Industry Minister Ed Husic said Washington’s move to sign an executive order governing the safe and responsible use of AI this week could help inform Australia’s response to the unregulated technology.

The move comes after Mr Biden issued a wide-ranging and ambitious executive order which AI experts say will catapult the US to become a front runner in regulating the technology.

The Australian government response comes ahead of the world’s biggest AI summit in Britain this week, hosted by British Prime Minister Rishi Sunak.

Australia will be represented by Mr Husic and Defence Minister Richard Marles, with the world leaders expected to discuss threats of AI and global efforts to regulate the technology.

The presidential executive order aims to address a range of areas including job losses, privacy concerns and consumer rights. It will also require companies developing AI to prove they are safe before rolling them out for wider use, with Mr Biden warning companies must “prove they pose no national security or safety risk to the American people”.

Mr Husic said Australia’s response would seek to “harmonise where we can, localise where we have to”.

The Australian understands the government is also following regulatory developments on AI in Indonesia and Singapore.

“We welcome this announcement,” Mr Husic said. “It signals serious intent by the US to get the balance right on the development, application and regulation of AI, especially generative AI.

“The US announcement can help to frame the path ahead for global co-operation as we gather to meet for the United Kingdom’s AI Safety Summit.”

The government is reviewing more than 500 submissions it has received to its consultation process to strengthen regulatory frameworks, identify gaps and ensure governance mechanisms are fit for purpose for AI.

The Australian understands Labor will release its response to the consultation process before the end of the year.

Australia’s AI strategy currently focuses on a broad set of general regulations, sector-specific regulation (airline, food and motor vehicle safety, therapeutic goods and financial services) and voluntary or self-regulation initiatives.

The AI Institute of UNSW’s chief scientist Toby Walsh said Mr Biden’s executive order was wide-ranging and ambitious, with the US having “put itself at the centre of the debate in terms of regulating AI”.

“The US has really seized the debate now, and the boy prime minister Rishi Sunak’s safety summit looks like a vacuous talk fest compared to the US signing an executive order to get things done,” Professor Walsh said.

“It‘s a welcome sign starting to see the US do that stuff, it has taken the US from zero to 100 miles an hour quite quickly.”

Professor Walsh said Australia should move quickly to regulate AI, warning the rise of artificial intelligence could sway voters in democratic elections in the next few years.

“We saw in the Slovakian election people possibly being moved by deep fakes, and half the world’s democracies get to vote next year,” he said. “We are about to supercharge that by putting AI-generated content and cloned audio, really believable stuff.

“We had this issue with money which is easy to fake but now have strict penalties for faking money … we need to do the same with counterfeit content, have severe penalties and police it rigorously.”

Mr Husic said the government was intent on promoting the safe and responsible use of AI.

“In Australia, we’ve been working with industry and listening to the community about how to promote the safe and responsible use of AI,” he said.

“When it comes to AI regulation, our country should aim to harmonise where we can, localise where we have to.”